K8S部署Hadoop集群

1、拉取镜像

到docker hub上拉取Hadoop镜像
[root@k8s-master hadoop]# docker pull kubeguide/hadoop:latest

#### 2、复制镜像到其他节点

```sh
# 打包镜像到本地
[root@k8s-master hadoop]# docker save e0af06208032 > hadoop.tar
# 发送到子节点
[root@k8s-master hadoop]# scp hadoop.tar k8s-node01:/root/hadoop
[root@k8s-master hadoop]# scp hadoop.tar k8s-node02:/root/hadoop
# 子节点构建镜像
[root@k8s-node01 hadoop]# docker load --input hadoop.tar
[root@k8s-node02 hadoop]# docker load --input hadoop.tar

3、编写hadoop.yaml

apiVersion: v1
kind: ConfigMap
metadata:
  name: kube-hadoop-conf
data:
  HDFS_MASTER_SERVICE: hadoop-hdfs-master
  HDOOP_YARN_MASTER: hadoop-yarn-master
---
apiVersion: v1
kind: Service
metadata:
  name: hadoop-hdfs-master
spec:
  type: NodePort
  selector:
    name: hdfs-master
  ports:
    - name: rpc
      port: 9000
      targetPort: 9000
    - name: http
      port: 50070
      targetPort: 50070
      nodePort: 32007
---
apiVersion: v1
kind: ReplicationController
metadata:
  name: hdfs-master
  labels:
    name: hdfs-master
spec:
  replicas: 1
  selector:
    name: hdfs-master
  template:
    metadata:
      labels:
        name: hdfs-master
    spec:
      containers:
        - name: hdfs-master
          image: kubeguide/hadoop:latest
          imagePullPolicy: IfNotPresent
          ports:
            - containerPort: 9000
            - containerPort: 50070
          env:
            - name: HADOOP_NODE_TYPE
              value: namenode
            - name: HDFS_MASTER_SERVICE
              valueFrom:
                configMapKeyRef:
                  name: kube-hadoop-conf
                  key: HDFS_MASTER_SERVICE
            - name: HDOOP_YARN_MASTER
              valueFrom:
                configMapKeyRef:
                  name: kube-hadoop-conf
                  key: HDOOP_YARN_MASTER
      restartPolicy: Always
---
apiVersion: v1
kind: ReplicationController
metadata:
  name: hadoop-datanode
  labels:
    app: hadoop-datanode
spec:
  replicas: 3
  selector:
    name: hadoop-datanode
  template:
    metadata:
      labels:
        name: hadoop-datanode
    spec:
      containers:
        - name: hadoop-datanode
          image: kubeguide/hadoop:latest
          imagePullPolicy: IfNotPresent
          ports:
            - containerPort: 9000
            - containerPort: 50070
          env:
            - name: HADOOP_NODE_TYPE
              value: datanode
            - name: HDFS_MASTER_SERVICE
              valueFrom:
                configMapKeyRef:
                  name: kube-hadoop-conf
                  key: HDFS_MASTER_SERVICE
            - name: HDOOP_YARN_MASTER
              valueFrom:
                configMapKeyRef:
                  name: kube-hadoop-conf
                  key: HDOOP_YARN_MASTER
      restartPolicy: Always
---
apiVersion: v1
kind: Service
metadata:
  name: hadoop-yarn-master
spec:
  type: NodePort
  selector:
    name: yarn-master
  ports:
     - name: "8030"
       port: 8030
     - name: "8031"
       port: 8031
     - name: "8032"
       port: 8032
     - name: http
       port: 8088
       targetPort: 8088
       nodePort: 32088
---
apiVersion: v1
kind: ReplicationController
metadata:
  name: yarn-master
  labels:
    name: yarn-master
spec:
  replicas: 1
  selector:
    name: yarn-master
  template:
    metadata:
      labels:
        name: yarn-master
    spec:
      containers:
        - name: yarn-master
          image: kubeguide/hadoop:latest
          imagePullPolicy: IfNotPresent
          ports:
            - containerPort: 9000
            - containerPort: 50070
          env:
            - name: HADOOP_NODE_TYPE
              value: resourceman
            - name: HDFS_MASTER_SERVICE
              valueFrom:
                configMapKeyRef:
                  name: kube-hadoop-conf
                  key: HDFS_MASTER_SERVICE
            - name: HDOOP_YARN_MASTER
              valueFrom:
                configMapKeyRef:
                  name: kube-hadoop-conf
                  key: HDOOP_YARN_MASTER
      restartPolicy: Always
---
apiVersion: v1
kind: Service
metadata:
  name: yarn-node
spec:
  clusterIP: None
  selector:
    name: yarn-node
  ports:
     - port: 8040
---
apiVersion: v1
kind: ReplicationController
metadata:
  name: yarn-node
  labels:
    name: yarn-node
spec:
  replicas: 3
  selector:
    name: yarn-node
  template:
    metadata:
      labels:
        name: yarn-node
    spec:
      containers:
        - name: yarn-node
          image: kubeguide/hadoop:latest
          imagePullPolicy: IfNotPresent
          ports:
            - containerPort: 8040
            - containerPort: 8041
            - containerPort: 8042
          env:
            - name: HADOOP_NODE_TYPE
              value: yarnnode
            - name: HDFS_MASTER_SERVICE
              valueFrom:
                configMapKeyRef:
                  name: kube-hadoop-conf
                  key: HDFS_MASTER_SERVICE
            - name: HDOOP_YARN_MASTER
              valueFrom:
                configMapKeyRef:
                  name: kube-hadoop-conf
                  key: HDOOP_YARN_MASTER
      restartPolicy: Always

这个yaml文件包含一个ConfigMap，5个Service，8个pod，这里需要注意的是ConfigMap中HDFS_MASTER_SERVICE和HDOOP_YARN_MASTER不要使用IP，使用HDFS service的名称，否则datanode将会连接不上namenode

4、执行创建命令

[root@k8s-master hadoop]# kubectl create -f hadoop.yaml

5、创建完成后就可以通过浏览器看到熟悉的HDFS管理界面了

【http://ip:32007】

【http://ip:32088】

备注(该方式未解决持久化存储问题)